3.1 Helix Score Diagnostic Version Pursuit Training
3.1.1 Helix Score
The ANOVA with repeated measures for the diagnostic version of the pursuit training revealed no significant interaction effect of group and time for the Helix Score (F(1, 31) = 1.11, p = .30, η²=.04; Table 2). No main effect of time (F(1, 31) = 2.20, p = .15, η²=.07), and group (F(1, 31) = 0.80, p = .43, η²=.03) could be found.
The ANCOVA with the covariate age revealed an almost significant interaction effect of time with the covariate age (F(1, 30) = 4.29, p = .05, η²=.13). Additionally there was a significant main effect of age (F(1, 30) = 12.17, p < .01, η²=.30). Based on this, a linear regression was calculated to examine the influence of age on the dependent variable (Helix Score (Pre and Post)). At the pre-measurement, age was found to be a significant predictor (β = − .02, t(32) = -2.55, p = .02). Age explained a significant proportion of the variance at the pre-measurement (R2 = .17, F(1,32) = 6.49, p = .02), the older the participants, the worse the Helix score. At the post-measurement, we no longer observed a significant effect (β < − .01, t(31) = − .39, p = .70; R2 = .07, F(1,31) = 0.15, p = .70).
3.1.2 Helix Score Reaction Time
For the ANOVA with repeated measures that compared the mean reaction time in the Helix-Arena between the groups, no significant interaction effect of group and time was found (F(1, 31) = 2.29, p = .14, η²=.07; Table 2). Furthermore, no significant main effect of time (F(1, 31) = 3.05, p = .09, η²=.09) or group (F(1, 31) = 0.01, p = .92, η²<.01) was observed.
The ANCOVA with the covariate age revealed no significant interaction effect of time with the covariate age (F(1, 30) = 0.02, p = .89, η²<. 01). But there was a significant main effect of age (F(1, 30) = 12.93, p < .01, η²=.30). The linear regression yielded age as a significant predictor at both the pre-measurement (β = .01, t(31) = 3.34, p < .01; R2 = .52, F(1,31) = 11.16, p < .01) and post-measurement performance (β = .01, t(31) = 2.57, p = .02; R2 = .42, F(1,31) = 6.61, p = .02). The older the participants, the higher the reaction time in the Helix-Arena.
Table 2
Descriptive statistics for the Helix Score and the reaction time in the Helix compared to the PC
| | Experimental group | | Control group |
| | N = 19 | | N = 14 |
Variable | | M | SD | | M | SD |
Helix-Score | Pre | 29.53 | .77 | | 29.57 | .76 |
Post | 29.95 | .23 | | 29.64 | .63 |
Helix(rt) | Pre | 2.21 | .51 | | 2.11 | .09 |
Post | 1.96 | .48 | | 2.09 | .44 |
Note. M = Mean. SD = Standard error; rt = reaction time. |
3.2 TAP
For the TAP, separate analyses were performed for each subtest
3.2.1 Alertness
For the subtest Alertness with the subcategories with or without sound, no significant interaction effect of group and time could be found (Table 3). Neither a main effect of time (with sound: F(1, 32) = 1.51, p = .23, η² = .05; without sound: F(1, 32) = 0.01, p = .94, η²<.01) or group (with sound: F(1, 32) = 0.39, p = .54, η² = .01; without sound: F(1, 32) = 0.35, p = .56, η² =.01) were significant.
The ANCOVA with the covariate age revealed no interaction effect of time and the covariate age (with sound: F(1, 31) = 0.10, p = .75, η² <.01; without sound: F(1, 31) = 1.45, p = .24, η²=.05). However there was a significant main effect of age (with sound: F(1, 31) = 7.61, p = .01, η² = .20; without sound: F(1, 31) = 15.91, p < .01, η² = .34). Therefore the linear regression showed significant results for age as a predictor at pre-measurement (with sound: β = 1.7, t(32) = 2.73, p = .01; R2 = .43, F(1,32) = 7.44, p = .01; without sound: β = 2.39, t(32) = 4.62, p < .01; R2 = .63, F(1,32) = 21.3, p < .01) and post-measurement (with sound: β = 1.49, t(32) = 2.52, p = .02; R2 = .40, F(1,32) = 6.37, p = .02; without sound: β = 1.68, t(32) = 2.81, p = .01; R2 = .45, F(1,32) = 7.89, p = .01). The older the participants, the higher the reaction time.
The ANOVA for the anticipate reaction parameter in the with sound subcategory found a significant interaction effect of group and time (F(1, 32) = 5.46, p = .03, η² = .15). At the pre-measurement, significantly more anticipated reactions were found in the EG than in the CG (t(32) = -2.99, p < .01 (CG: M = 0, SD = 0; EG: M = .47, SD = .61). At post-measurement, this difference could no longer be detected (t(32) = .07, p = .47 (CG: M = .33, SD = .62; EG: M = .32, SD = .75). In the CG, a significant decrease from pre- to post-measurement was found (t(14) = -2.09, p = .03). No significant change was found in the EG ( t(18) = 1.14, p = .13). There was no main effect of time (F(1, 32) = 0.70, p = .41, η²= .02) and group (F(1, 32) = 1.72, p = .20, η²= .05).
3.2.2 Shared Attention
In addition, the subtest Shared Attention with the subcategories auditory and visual was applied. No significant interaction effect of group and time was found, neither for auditory nor for visual (Table 3).
For the auditory subtest, neither a significant main effect of time (F(1, 32) = 2.13, p = .15, η²= .06) and group (F(1, 32) = 1.85, p = .18, η²= .06), could be found.
An interaction effect of time with the covariate age could not be found (F(1, 31) < 0.01, p = .93, η²<.01). Nevertheless, there was a significant main effect of age (F(1, 31) = 8.18, p < .01, η²= .21. The linear regression revealed age as a significant predictor for pre-measurement (β = 2.3, t(32) = 2.69, p = .01; R2 = .43, F(1,32) = 7.23, p = .01) and post-measurement (β = 2.20, t(32) = 2.98, p < .01; R2 = .47, F(1,32) = 8.89, p < .01). The older the participants, the worse the reaction time.
The ANOVA for the auditory subtest with regard to the outlier parameter revealed no significant interaction effect of group and time (F(1,32) = 2.17, p = .15, η²=.06). In addition, no main effect of time (F(1,32) = 0.23, p = .63, η²<.01) and group were found (F(1,32) = 1.17, p = .29, η²=.04).
Whereas in the visual subtest no significant main effect of time (F(1, 32) = 3.48, p = .07, η²=.10), and group (F(1, 32) = 0.90, p = .35, η² = .03). could be found.
The ANCOVA with the covariate age revealed an interaction effect of age and time (F(1, 31) = 7.99, p = .01, η²=.21). Additionally there was a significant main effect of age (F(1, 31) = 46.67, p < .01, η²= .60). The further results for the interaction effect with age as well as the main effect of age revealed that there were significant linear regressions at both the pre-measurement (β = 3.91, t(32) = 4.84, p < .01; R2 = .65, F(1,32) = 23.45, p < .01) and the post-measurement (β = 5.85, t(32) = 8.15, p < .01; R2 = .82, F(1,32) = 66.45, p < .01). That is, the older the participants, the worse the reaction time in the visual divided attention task.
The ANOVA for the visual subtest with regard to the outlier parameter revealed no significant interaction effect of group and time (F(1,32) = 1.04, p = .32, η²=.03). In addition, no main effect of time (F(1,32) < 0.01, p = .98, η²<.01) and group were found (F(1,32) = 2.36, p = .13, η²=.07).
For both subcategories auditory and visual no significant differences could be found in the ANOVA with the common parameter error. No significant interaction effect of group and time could be found (F(1, 32) = 0.31, p = .58, η²=.01). No main effect of time (F(1, 32) = 0.31, p = .58, η²=.01), and no main effect of group (F(1, 32) = 0.82, p = .37, η²=.03) were found.
3.2.3 Attention Shift
For Attention Shift the subcategories valid instructor and invalid instructor were analyzed. For the valid instructor no significant interaction effect of group and time could be found (Table 3).
However, a significant interaction effect of group and time was found for the invalid instructor (F(1,32) = 4.45, p = .04, η²=.12). Further analyses of the interaction effect of the time (invalid instructor) and the group revealed a significant improvement in reaction time between pre- and post-measurement. This effect was found in the EG (t(18)=-2.9, p <. 01), but not in the CG (t(14)=--.25, p = .40; Table 3). EG participants were faster at post-measurement than pre-measurement.
No significant results were observed for the main effect time (valid: F(1, 32) = 1.20, p = .28, η²=.04; invalid: F(1, 32) = 3.02, p = .09, η² =.09) or group (valid: F(1, 32) = 1.50, p = .23, η²=.05; invalid: F(1, 32) = 1.29, p = .26, η² =.04).
The ANCOVA with the covariate age revealed no significant interaction effect of the covariate age and time (valid: F(1, 31) = 0.17, p = .69, η²=.01; invalid: F(1, 31) = 1.29, p = .27, η²=.04). However there was a significant main effect of age (valid: F(1, 31) = 27.74, p < .01, η²=.47; invalid: F(1, 31) = 34.77, p < .01, η²=.53). The further calculations revealed significant linear regressions for age as a predictor at pre-measurement (valid: β = 3.09, t(32) = 5.31, p < .01; R2 = .68, F(1,32) = 28.16, p < .01; invalid: β = 3.89, t(32) = 5.93, p < .01; R2 = .72, F(1,32) = 35.11, p < .01) and post-measurement (valid: β = 2.82, t(32) = 4.82, p < .01; R2 = .65, F(1,32) = 23.22, p < .01; invalid: β = 3.15, t(32) = 5.12, p < .01; R2 = .67, F(1,32) = 26.18, p < .01). The older the participants, the worse the reaction time in both the valid- and invalid-instructor of the attention shift task.
For both subcategories valid and invalid no significant differences could be found at the ANOVA with the common parameter anticipate reactions. No significant interaction effect of group and time could be found (F(1, 32) = 0.67, p = .42, η²=.02). No main effect of group (F(1, 31) = 0.42, p = .52, η²=.01), was found. Nevertheless there was a main effect of time (F(1, 32) = 4.89, p = .03, η²=.13). There were more anticipate reactions in the pre- than the post-measurement (Pre: M = 2.34,SD = 3.48; Post: M = 1.24, SD = 1.10).
3.2.4 Visual Scanning
The last subtest of the TAP that was applied was Visual Scanning. The subcategories critical and non critical were examined. No significant interaction effect of group and time was found, for both critical and non-critical value (Table 3).
With regard to the main effect of time, a significant result was found for the critical value (F(1, 30) = 11.28, p < .01, η²=.27). The total sample was faster at post-measurement (M = 2231.63; SD = 700.96) than at pre-measurement (M = 2499.53; SD = 954.72). There was no significant main effect of group (F(1, 30) = 0.73, p = .40, η²=.02).
For the non-critical value no significant main effect of time was found (F(1, 30) = 3.51, p = .07, η²=.11). The same occurred for the main effect of group, no significant result was found (F(1, 30) = 0.16, p = .96, η²=.01).
The ANCOVA with the covariate age revealed a non significant interaction effect of the covariate age and time for the critical value (F(1, 29) = 3.94, p = .06, η²=.12). Further there was a significant main effect of age (F(1, 29) = 21.26, p < .01, η²=.42). For both pre- and post-measurement, a significant linear regression with age as a predictor was determined. Older participants had a slower reaction time at pre-measurement (β = 38.06, t(31) = 4.57, p < .01; R2 = .63, F(1,31) = 20.86, p < .01), and post-measurement (β = 28.80, t(31) = 4.72, p < .01; R2 = .65, F(1,31) = 22.32, p < .01). The interaction-effect with age (F(1, 29) = 6.81, p = .01, η²=.19), and the main effect of age (F(1, 29) = 16.68, p < .01, η²=.37), was also found for the non-critical value. To pre-measurement (β = 79.59, t(31) = 4.09, p < .01; R2 = .59, F(1,31) = 16.72, p < .01), and post-measurement (β = 53.04, t(30) = 3.93, p < .01; R2 = .58, F(1,30) = 15.48, p < .01), older participants were slower.
For the subcategory critical value and the parameter outlier, the ANOVA revealed a significant interaction effect of group and time (F(1, 30) = 4.02, p = .05, η²=.12). For the pre-measurement, a significant difference between the groups was found (t(31) = 1.75, p = .05; CG: M = 7.57, SD = 6.21; EG M = 4.33, SD = 4.04). No significant difference was found for post-measurement (t(30) = .12, p = .45; CG: M = 4.29, SD = 4.41; EG: M = 4.11, SD = 3.92). In the CG, a significant improvement from pre- to post-measurement was found (t(13) = 2.22, p = .02). Not in the EG (t(17) = .31, p = .38). Further there was a significant main effect of time (F(1, 30) = 5.27, p = .03, η²=.15). There were more outlier in the pre- than the post-measurement (Pre: M = 5.75,SD = 5.27; Post: M = 4.19, SD = 4.08). There was no significant main effect of group (F(1, 30) = 1.35, p = .25, η²=.04).
With regard to the error parameter in the non critical subtest, the ANOVA revealed no significant interaction effect of group and time (F(1, 30) < 0.01, p = 1, η²<.01). No main effect of time (F(1, 30) < 0.01, p = 1, η²<.01) and no main effect of group (F(1, 30) = 2.28, p = .14, η²=.07) were found.
Table 3
Results of the TAP subtests, (Mean reaction time (ms); SD in Brackets)
Variable | | |
Experimental group N = 19 | Control group N = 15 | F-value (df) (group*time), p-value |
Alertness | |
without sound | Pre | 270.05 (69.85) | 249.27 (44.24) | F(1,32) = 1.30, p = .26 |
Post | 260.00 (50.03) | 258.13 (72.43) |
with sound | Pre | 277.16 (72.17) | 259.13 (46.68) | F(1,32) = 0.60, p = .44 |
Post | 261.48 (55.65) | 255.67 (62.53) |
Shared Attention | |
Auditory | Pre | 603.58 (85.32) | 560.53 (83.75) | F(1,32) = 0.68, p = .42 |
Post | 581.79 (72.40) | 554.47 (77.62) |
Visual | Pre | 794.37 (91.55) | 775.40 (101.91) | F(1,32) = 1.43, p = .24 |
Post | 786.79 (123.77) | 740.80 (95.57) |
Attention Shift | | | |
valid instructor | Pre | 313.95 (79.99) | 275.20 (54.45) | F(1,32) = 2.40, p = .13 |
Post | 296.00 (69.10) | 278.27 (69.70) |
invalid instructor | Pre | 352.63 (91.08) | 305.60 (71.839 | F(1,32) = 4.45, p = .04 |
Post | 321.00 (73.53) | 308.67 (77.37) |
Visual Scanning | | | |
critical | Pre | 2597.11 (621.90) | 2348.79 (1266.77) | F(1,30) = 0.08, p = .79 |
Post | 2330.00 (581.86) | 2105.14 (835.66) |
non critical | Pre | 4711.79 (1120.59) | 4534.50 (3029.03) | F(1,30) = 0.40, p = .84 |
Post | 4439.06 (1026.00) | 4147.92 (1885.53) |
3.3 PANAS
The ANOVA with repeated measures for the PANAS results were calculated separately for each scale (positive and negative).
3.3.1 Positive Subscale
The positive scale revealed a significant interaction effect of group and time (F(1, 32) = 14.86, p < .01, η²=.32). It shows a higher score in the EG compared to the CG (t(32) = -3.09, p < .01; Table 4, Fig. 4) at post-measurement. At pre-measurement no significant difference could be observed (t(32) = -1.22, p = .12). In addition, a significantly higher PANAS score was found in the EG for post-measurement than for pre-measurement (t(18) = -5.68, p < .01; Table 4). This effect was not found for the CG (t(14) = − .5, p = .31). Additionally there was a main effect of group (F(1, 32) = 5.14, p = .03, η²=.14) and time (F(1, 32) = 20.09, p < .01, η²=.39). The positive PANAS Score was in the post-measurement (M = 3.60, SD = 0.88) higher than in the pre-measurement (M = 3.26,SD = 0.74) .
The ANCOVA with age, revealed no interaction effect of the covariate age and time (F(1, 31) = 0.06, p = .81, η²<.01). However, there was a significant main effect of age (F(1, 31) = 4.76, p = .04, η²=.13). Further tests revealed a significant linear regression with age as a predictor for pre-measurement (β = .02, t(32) = 2.28, p = .03; R2 = .37, F(1,32) = 5.18, p = .03) and post-measurement (β = .02, t(32) = 2.28, p = .03; R2 = .37, F(1,32) = 5.22, p = .03). The older the participants, the higher the score of the positive PANAS subscale.
Table 4
Descriptive statistics PANAS
| | Experimental group | | Control group |
| | N = 19 | | N = 15 |
Variable | | M | SD | | M | SD |
Positive affect | Pre | 3.39 | 0.73 | | 3.08 | 0.74 |
Post | 3.97 | 0.71 | | 3.13 | 0.88 |
Negative affect | Pre | 1.10 | 0.17 | | 1.22 | 0.31 |
Post | 1.07 | 0.13 | | 1.13 | 0.26 |
Note. M = mean. SD = standard deviation |
3.3.2 Negative Subscale
The negative scale revealed no significant interaction effect of group and time (F(1, 32) = 2.16, p = .15; Table 4). Moreover, there was a significant main effect of time (F(1, 32) = 7.87, p < .01, η² = .20). The negative PANAS Score was in the post-measurement (M = 1.10, SD = 0.20) lower than in the pre-measurement (M = 1.15,SD = 0.25).
No significant main effect of group (F(1, 32) = 1.42, p = .24, η² = .04), could be found.
The ANCOVA with age revealed no significant interaction effect with age (F(1, 31) = 0.03, p = .87, η²<.01), and no main effect of age (F(1, 31) = 3.71, p = .06, η² = .11).